During the manufacture of microelectronic devices, such as memory chips, processor chips and field emission displays, etching processes are often used to form features on a microelectronic substrate or substrate assembly that forms the foundation of the device. A typical etching technique includes depositing a layer of a photoresist material on the substrate, masking selected portions of the layer and exposing the unmasked portions to a selected radiation. The selected radiation changes the solubility of the unmasked portions to become either soluble (in the case of a positive photoresist) or insoluble (in the case of a negative photoresist) when exposed to a selected solvent. The photoresist layer is then washed with the selected solvent to remove either the exposed or unexposed photoresist material, exposing a portion of the substrate beneath. The substrate is washed with an etchant that removes material from the exposed portions of the substrate while leaving intact the portions of the substrate covered by the photoresist material.
It is often important to control the uniformity of the thickness to which the photoresist material is deposited on the substrate. For example, if the photoresist material is deposited to a nonuniform thickness, certain portions of the photoresist material may be overexposed to the radiation while other portions may be underexposed. Where the photoresist material is overexposed, the edges between the masked and unmasked regions can become blurred, making the process unsuitable for forming very, small features. Where the photoresist material is underexposed, it may not have sufficient exposure time to change solubility. Furthermore, it may be desirable to keep the overall thickness of the photoresistant layer relatively small to increase the resolution of the features formed with this technique.
The photoresist material is typically deposited on the substrate or substrate assembly by disposing the material in liquid form at the center of the substrate and spinning the substrate about its center to spread the material outwardly by centrifugal force. One drawback with this technique is that the liquid photoresist material can interact with the adjacent air mass, creating waves or other disturbances in the photoresist material that affect the uniformity of the layer thickness. This problem can become more acute when the velocity of the substrate increases, for example, when the substrate is rotated at a high angular velocity and/or when the substrate has a large radius so that at even moderate angular velocities, the linear speed toward the edge of the substrate is high.
Another drawback with this technique is that the convective heat transfer rate can vary over the surface of the substrate because the relative linear velocity between the substrate and the adjacent air mass varies with the distance from the substrate center. The variation in heat transfer rates can cause the surface temperature of the substrate to vary, in turn causing the evaporation rate of the fluid (and therefore the thickness of the fluid) to vary over surface of the substrate.
Yet another drawback with this technique is that the viscosity selected for the liquid photoresist material must account for the diameter and rotation speed of the substrate. For example, a relatively viscous liquid may be selected for large substrates to prevent the liquid from flying off the edges of the substrate before accumulating to the desired thickness. Such a liquid may be too viscous for smaller substrates. Accordingly, conventional techniques typically use liquids with different viscosities to form layers having different thicknesses. For example, less viscous liquids can be used to form thinner layers and more viscous liquids can be used to form thicker layers. One problem with this approach is that it requires controlling and/or adjusting the viscosity of the liquid and/or providing multiple sources of the liquid, each having a different viscosity. Furthermore, while the angular velocity of the substrate can be used to control the thickness of the liquid layer (for example, by increasing the angular velocity to reduce the layer thickness), this technique is limited because at high angular velocities, the liquid can form waves or other disturbances, as discussed above.
FIG. 1 is a partially schematic, partially cutaway side elevation view of a conventional device 10 that can address some of the foregoing problems for rectangular substrates. The device 10 includes a motor 30 having a shaft 32 connected to a chuck 33 and a bowl 20. A substrate 12 having a rectangular planform shape is releasably mounted to the chuck 33 and both the substrate 12 and the bowl 20 spin as the shaft 32 rotates. Accordingly, the air adjacent to the substrate 12 is partially contained within the spinning bowl 20 so that at least a portion of the air will spin at the same rate as the substrate 12. A fluid supply conduit 23 disposes a liquid onto the substrate 12 through an aperture 24 and the liquid spreads out over the surface of the substrate 12 as the substrate 12 spins. Excess liquid is collected in the bowl 20 as it runs over the edges of the substrate 12 and can be removed from the bowl via a drain 21. Air can be exhausted from the bowl 20 through an exhaust port 22.
One potential drawback with the device 10 shown in FIG. 1 is that the bowl 20 can be heavy and difficult to spin smoothly at high rates of speed. Furthermore, the drain 21 and the exhaust port 22 may be coupled to a drain line 23a and an exhaust line 23b, respectively, which must be secured to the bowl 20 with fluid-tight rotating couplings. Still further, the bowl 20 is partially open so that it may be time consuming to bring the air mass adjacent to the substrate 12 up to the same rotational speed as the substrate 12, particularly where the substrate 12 rotates at high speed.
FIG. 2 is a partially schematic, partially cutaway side elevation view of another conventional device 10a that includes a motor 30a coupled with a shaft 32a to a chuck 33a. The chuck 33a includes a rectangular recess 36 for receiving the rectangular substrate 12. A cover 40 is releasably placed on the chuck 33a to rotate with the chuck 33a and the substrate 12. The cover 40 includes an aperture 41 that allows fluid to pass from the fluid supply conduit 23 to the surface of the substrate 12. The apparatus 10a can further include a collection vessel 20a fixed relative to the motor 30a and having a drain 21 and an exhaust port 22 for removing liquid and gas from the region adjacent to the substrate 12.
One problem with the device 10a shown in FIG. 2 is that the liquid disposed on the substrate 12 can become trapped between the lower surface of the substrate 12 and the walls of the recess 36 into which the substrate 12 is placed. A further drawback is that the recess 36 is sized for rectangular substrates 12, making it unsuitable for or unusable with round substrates, particularly where the diameter of the round substrate exceeds the width of the recess 36.